Download Competitive abilities of introduced and native grasses

(
.,
117
Plant Ecology 157: 117-125, 2001.
© 2001 Kluwer Academic Publishers . Printed in the Netherlands.
Competitive abilities of introduced and native grasses
Jonathan Bakker-' and Scott Wilson 1.*
'Department of Biology, University of Regina, S4S OA2, Regina, Saskatchewan, Canada; 2TRP Forestry
Consultants, 2108 23rd Avenue, Vernon, VIT lJ4, British Columbia, Canada; *Author for correspondence
(e-mail: scott.wilsontisleroy.cc.uregina.ca; fax: 306-585-4894, phone: 306-585-4287)
Received 7 December 1999; accepted in revised form 10 April 2000
G'Y
words: Agropyron cristatum, Bouteloua gracilis Competition, Mixed-grass prairie, W'''')
Abstract
Differences in competitive ability may explain the maintenance of existing plant populations and the invasion of
new areas by plant species. We used field experiments to examine the competitive responses of Agropyron cris­
tatum (L.) Gaertn., an introduced C 3 grass, and Bouteloua gracilis (HBK.) Lag., a native C 4 grass, and the com­
petitive effects of Agropyron-dominated vegetation and successional prairie. We also tested whether the outcome
of competitive interactions varied with water availability. In each vegetation type. transplants of each species
were grown under two levels of competition (presence or absence of neighboring vegetation) and three levels of
water availability (high. medium, or low). Transplant survival, growth. and biomass allocation patterns were
measured. Water availability had no effect on the measured variables, suggesting that both species were limited
by another resource. Growth rates were affected more by competition, while survival and root: shoot ratio were
affected more by transplant species identity. In the successional prairie, neighboring vegetation suppressed the
growth of Agropyron transplants less than that of Bouteloua transplants, suggesting that Agropyron has a stronger
ability to resist competitive suppression in that vegetation type. The spread of Agropyron into surrounding veg­
etation may relate to its ability to resist competitive suppression. In the Agropyron-dominated vegetation, neigh­
boring vegetation suppressed the growth of both species by the same extent. However, competition accounted for
more variation in transplant growth in Agropyron-dominated vegetation than in successional prairie, suggesting
that Agropyron has strong competitive effects which hinder plant growth and prevent other species froin estab­
lishing in Agropyron fields.
Introduction
Human activity has altered the distribution and abun­
dance of many species (Drake et al. 1989). Once es­
tablished in new areas, introduced species can alter
ecosystem structure and function, and reduce habitat
for native plants and animals (Vitousek 1990).
Introduced grasses are especially adept invaders of
North American grasslands (D' Antonio and Vitousek
1992). Agropyron cristatum (L.) Gaertn., an Asian C 3
grass. has been seeded throughout western North
America because it establishes rapidly, prevents ero­
sion. provides nutritious spring forage (Lesica and
DeLuca 1996), and produces more forage than native
grasses (Lawrence and Ratzlaff 1989). However,
Agropyron also limits successional changes Looman
and Heinrichs (1973). invades adjacent lands (Hull
and Klomp 1967) , reduces plant and animal diversity
(Wilson 1989; Sutter and Brigham 1998), and alters
soil chemistry (Donnaar et a!. 1995; Chri stian and
Wilson 1999).
Differences in competitive ability may explain the
maintenance of existing plant populations and the in­
vasion of new areas by plant species . Competitive
ability includes the effect of competition from an in­
dividual on neighboring individuals (competitive ef­
fect) and the response of an individual to competition
from neighboring individuals (competitive response;
Goldberg (1990». The paucity of native species from
Agropyron fields (Looman and Heinrichs 1973; Wil­
son 1989; Christian and Wilson 1999) suggests that
Agropyron is a strong competitor which can displace
118
or prevent the establishment of native species. How­
ever, Agropyron also responds to competition: neigh­
boring Bromus inermis Leyss . suppressed Agropyron
transplants to about the same extent as it suppressed
transplants of Bouteloua gracilis (HBK .) Lag., a C 4
grass native to the northern Great Plains of North
America (Gerry and Wilson 1995).
We hypothesized 'that introduced and native
grasses would differ in competitive ability and that
these differences would be evident in their growth and
biomass allocation patterns. To test this hypothesis,
we used field experiments to examine the competitive
responses of Agropyron and Bouteloua transplants
and the competitive effects of Agropyron-dominated
vegetation and successional prairie. Transplants were
used to avoid interspecific differences in germination
and establishment.
In environments with limited productivity, most
competition occurs for below ground resources such
as water and mineral nutrients (Casper and Jackson
1997). In the northern Great Plains , primary produc­
tion is strongly correlated with precipitation (Sala et
al. 1988), suggesting that water is a limiting resource
in this region. In addition, interspecific differences in
water use efficiency, ontogenetic patterns (Morgan et
al. 1998), ' and response to summer drought (Frank
1994) suggest that species should respond differently
to water availability. We hypothesized that water
availability would affect the outcome of competitive
interactions. We used high, medium, and low levels
of water availability to test this hypothesis.
Methods
Study site
Grasslands National Park (49°22'N 107°53'W) is lo­
cated in the mixed-grass prairie region of southern
Saskatchewan, Canada. The vegetation is dominated
by grasses tBouteloua, Koeleria gracilis Pers., Stipa
comata Trin . & Rupr.) and spikemoss (Selaginella
densa Rydb. ; nomenclature follows Barkworth and
Dewey (1985) for the Triticaceae and Looman and
Best (1987) for all other taxa). The region has a con­
tinental climate with average temperatures of - 7 °C
in January and 27°C in July . The average annual pre­
cipitation is 313 nun, 46% of which occurs from May
through July (Environment Canada 1966--1993).
Mean temperatures and precipitation levels experi­
enced during the experiment were within one stan­
dard deviation of the long-term (1966-1993) mean.
The experiment was conducted in a field that had
been cultivated and abandoned around 1940. After
abandonment, part of the field was sown with Agro­
pyron and part of it underwent natural succession to
dominance by native grasses. As a result, the field
contained two vegetation types (Agropyron-dominat­
ed vegetation and succession al prairie ) with similar
disturbance histories, slopes, and aspects . Soils in the
field were loam to clay loams .
Experimental design
We used root exclusion tubes to restrict the competi­
tive interactions of the transplant to a defined area.
Each tube (pVC, 10 cm diameter, 15 em length) was
hammered vertically into the ground until its upper
edge was I em above the soil surface. Tub es were ar­
ranged in a grid with 3 m spacing. Transplants were
much smaller (final mass < I g) than the soil mass
within each tube ( ~ 1 kg).
Vegetation types were not interspersed, so a sepa­
rate factorial experiment was performed in each type .
The experimental design consi sted of 2 competition
levels x 3 water availability levels x 2 transplant spe ­
cies. Treatment combinations were replicated 10
times for a total of 120 tubes per vegetation type.
Treatments were randomly assigned to tubes.
Two competition levels were used: AN, with all
neighbors present, and NN, with no neighbors
present. Vegetation in AN tubes was not disturbed.
Vegetation in NN tubes was sprayed with a 1% gly­
phosate [N-(phosphonomethyJ )glycine] solution two
weeks before seedlings were transplanted. Glyphosate
degrades rapidly and has no residual activity. Re­
growth of neighboring vegetation was continually re­
moved by hand.
High, medium, and low levels of water availabil­
ity were calculated as the monthly means of the 4
wettest, 4 median, and 4 driest summers (May to Au­
gust) in the last 30 years at Val Marie, Saskatchewan
(about 10 km north of the study site ; Environment
Canada (1966-1993». Water was applied biweekly
from May to August 1995 using historical monthly
precipitation patterns (Figure I). Natural precipitation
was reduced by positioning a 20 x 20 cm piece of
colorless Acrylite FF (Johnson Plastics, Winnipeg,
Manitoba, Canada) 15 em above each tube. Acrylite
. FF is an acrylic material that transmits 92% of light
and 89% of solar energy.
119
100
E
.s- eo
C
:J
0
60
E
'"
.~
'"
~
40
20
0
May
June
July
August
Month
ing vegetation were cut and separated, and then each
tube, with the soil mass encompassed by it, was re­
moved from the ground . After soaking each soil mass,
the transplant root system was carefully teased, in its
entirety, from the soil. All remaining roots were as­
sumed to be from neighboring vegetation within the
tubes. Root samples were washed to remove all soil
from the surface. All vegetation samples were dried
to constant mass and weighed .
The relative growth rate (R) of each transplant was
calculated as:
Figure J. High (squares), medium (circles), and low (triangles)
levels of water availability under which transplants were grown.
Agropyron and Bouteloua seeds were sown in
April 1995 into pots (2.5 ern diameter, 12 em deep)
filled with a 3:1 mixture of peat and local topsoil. Af­
ter germination, densities were reduced to one indi­
vidual per pot. Seedlings were grown in a greenhouse
at the University of Regina, where they received nat­
ural light and daily water. After 5 weeks, seedlings
were moved outside for a week of hardening. In late
May 1995, seedlings were brought to the study site
where one individual was transplanted into the center
of each tube. All transplants received 31 mm of water
during transplanting and 28 mm a week later to aid
establishment, and were shaded for 2 weeks after
transplanting "to reduce desiccation, Transplants that
died within 2 weeks of planting were replaced.
Sampling methodology and analysis
Initial total transplant mass (M j ; dry weight in g) was
calculated using species-specific regressions (Gerry
and Wilson 1995) with total tiller length (ITL; mm):
Agropyron:M,= [0.OOO859(ITL)-0.OO 174 ]2;r2
=0.91;n=1O
Bouteloua:M i=[0 .OOO796(1TL)+0.00458]2;r2
=0.98;n=1O
TIL was defined as the sum of the lengths of the
longest leaf of each tiller of a transplant and was
measured before transplanting. The initial transplant
mass of both species was 0.024 g. The initial root:
shoot ratios of Agropyron and Bouteloua were 0.99
and 1.15, respectively.
Transplants were harvested in mid-September
1995. Transplant shoots and aboveground neighbor­
where M, is final total mass, M ; is initial total
mass, and d is the number of days between measur­
ing 1TL and harvest.
Final transplant root: shoot ratios (R:S) were cal­
culated as:
R:S=M/M,
where M, is final root mass and M. is final shoot
mass.
Statistical analyses were performed using JMP sta­
tistical software (version 3.0.2; SAS Institute (1994».
Vegetation types were analyzed separately. Log-like­
lihood ratios tested the effects of competition, water,
and species on transplant survival to harvest. Three­
way analyses of variance (ANOVAs) tested the ef­
fects of competition, water, species, and all interac­
tions on growth rate and on root: shoot ratio.
To further examine the effect of competition on
" transpla nt performance, transplant growth (i.e.,
MIMi) was regressed on aboveground neighbor
mass in AN tubes. Aboveground neighbor masses
were log transformed to improve the model fit. Lin­
ear equations were fit to significant (P = 0.05) regres­
sions. The 95% confidence limits of the regressions
were compared to detect significant differences
among regressions. For each equation, R 2 estimates
the proportion of variation in transplant growth that
can be accounted for by aboveground neighbor mass,
the slope measures the per-gram effect of neighbors
on transplant growth (Goldberg 1987), and the y-in­
tercept predicts transplant growth in the absence of
neighbors.
120
Agropyron-dominaled
Success ional prairie
AgropyrotHlomlnated
vegetation
100
s
0,03
e
OJ
80
.t::_
e
..c;i
60
~ ·a
a.
<10
g.~
z
::l
j"'c
2:-
i'!
I­
0.02
:ijS
lO­
l!!
Successional prairie
vegetation
c:
i'!
0.01
I­
2Q
0.00
AN
I'll'l
Agropyron
AN
NN
Bout&loua
AN
lIN
Agropyron
AN
NN
Bouteloua
Figure 2. Mean survival of Agropyron and Bouteloua transplants
in Agropyron-dominated vegetation and successional prairie grown
with all neighbors (AN) or no neighbors (NN) and under low
(open ), medium (shaded), or high (dark) levels of water availabil­
ity. Log-likelihood ratio results in (Table I).
AN
NN
A{;nlpyron
AN
NN
Bout.....
AgropyrOlHiominated
Results
Competition experiments
Variation in water availability had no effect on trans­
plant survival (Figure 2), growth rate (Figure 3), or
. root: shoot ratio (Figure 4) in either vegetation type.
Further, no interactions with water were significant
(Table I). .
AN
NN
BoU.1oua
Figure 3. Mean relative growth rate (±SE) of Agropyron and
Bouteloua transplants in Agropyron-dominaled vegetation and suc­
cessional prairie grown with all neighbors (AN) or no neighbors
(NN) and under low (open), medium (shaded) , or high (dark) lev­
els of water availability . ANOVA results in (Table I).
Successional prairie
vegelation
Comparison of vegetation types
The species composition of each vegetation type was
assessed in late July 1995 using eight 0.5 x 1.0 m
quadrats distributed around the perimeter of each
study site. In each quadrat, the percent cover of bare
ground, litter, and each plant species was estimated
using Daubenmire's scale (Mueller-Dombois and El­
lenberg 1974). The 0-5% class was broken into 0-1 %
and 1-5% classes to more accurately reflect the im­
portance of smaller and infrequent plants. Species
richness was calculated for each quadrat.
Shoot mass was sampled by clipping all vegetation
1 cm above the soil surface in a 0.1 m 2 area at each
quadrat. Live shoots (the current year's growth) were
separated from dead shoots. Root mass was sampled
by pooling three soil cores (2 cm diameter, 10 cm
depth) per quadrat and washing them to separate roots
from soil. Live and dead roots could not be separated.
Shoot and root samples were dried to constant mass
and weighed . Root: shoot ratios were calculated for
each vegetation type.
Differences between vegetation types were tested
using Kruskal-Wallis one-way ANOVAs on ranks.
AN
NN
Agrq>yron
AN
NN
A~
BouI_
AN
NN
AN
NN
Agnpyron
AN
NN
BoutoA:>ue
Figure 4. Mean root: shoot ratio (±SE) of Agropyron and Boutel­
oua transplants in Agropyron-dominaled vegetat ion and succes­
sional prairie grown with all neighbors (AN) or no neighbors (NN)
and under low (open), medium (shaded ), or high (dark ) levels of
water availability. ANOVA results in (Table I).
Transplant survival was significantly lower for
Agropyron than Boute/oua in both vegetation types
(Figure 2). Survival tended (P ::: 0.056) to be higher
when neighbors were present in the Agropyron-domi­
nated vegetation but was not affected by the presence
or absence of neighbors in the successional prairie.
Interactions were not tested as lack-of-fit tests indi­
cated that including them would not improve the
model fit.
Transplant growth rate was significantly lower in
the presence 'of neighbors in both vegetation types
(Figure 3). In the Agropyron-dominated vegetation,
transplant growth rate was significantly higher for
Bouteloua than Agropyron and the competition x spe­
cies interaction was not significant. In the succes­
sional prairie, the presence of neighbors had a larger
effect on the growth rate of Bouteloua than Agropy­
ron transplants, producing a significant competition x
species interaction. Overall, transplant growth rates
were higher in the Agropyron-dominated vegetation
than the successional prairie.
121
Table I . Log-likelihood ratio results for transplant surv ival (Figure 2) and ANOVA results for relative growth rate (Figure 3) and root : shoot
ratio (Figure 4) in Agropyron.dominated vegetation and successional prairie. C : competition effect; W: water effect; S : species effect. ' : P <
0.05; " : P < 0.01 ; "': P < 0.001.
Source
Survi val
df
G
Root: shoot ratio
Relative growth rate
MS
F-R_tio
df
df
MS
F-Ratio
7.52"
Agropyron-dominated vegetation
C
I
3.663
1
36.428
95.29'"
I
0.702
W
2
2
0.682
1.78
2
0.131
1.40
S
CxW
1
4.220
27.573'­
I
2.564
6 .71'
1
3.488
37.34'"
CxS
2
0.700
1.83
2
0.132
1.42
1
0.580
1.52
1
0.190
2.03
Wx S
2
0.404
1.06
2
0.155
1.66
C x Wx S
2
0.899
2.35
2
0.084
0.90
Error
84
0.382
84
0.093
Successional prairie
I
2.246
1
27.317
45.49'"
1
0.474
3.92'
2
1.536
2
0.050
0.08
2
0.100
0.83
1
18.621'"
1
0.565
0.94
1
6.850
56.75'"
2
0.627
1.04
2
CUOS
0.87
CxS
I
3.523
5.87'
I
0.120
0.99
Wx S
2
0.095
0.16
2
0.031
0.26
Cx Wx S
2
0.282
0.47
2
0.057
0.47
Error
8S
0.601
83
0.121
C
W
S
CxW
Transplant root shoot ratios in both vegetation
types were significantly higher in the presence of
neighbors and significantly lower for Bouteloua than
Agropyron (Figure 4). Competition x species interac­
tions were not significant in either vegetation type.
Overall, root: shoot ratios were higher in the succes­
sional prairie than the Agropyron.dominated vegeta­
tion.
Transplant growth was significantly negatively
correlated with aboveground neighbor mass for both
species in both vegetation types (Figure 5). Regres­
sion equation s did not differ from one another in slope
or y-intercept, Aboveground neighbor mass explained
up to 40% of the variation in transplant growth, but
explained much less of the variation in Bouteloua
than Agropyron growth (Figure 5: R2 values). Per­
gram effects tended to be lower for Bouteloua than
Agropyron (Figure 5: slopes). Predicted and actual
growth in the absence of neighbors were similar for
Agropyron (Figure 5: y-intercepts), Predicted growth
was much lower than actual growth in the absence of
neighbors for Bouteloua.
Comparison of vegetation types
Species richness was significantly higher in succes­
sional prairie than Agropyron-dominated vegetation
(Table 2). Agropyron cover was significantly higher
in Agropyron-dominated vegetation. Covers of all
other plants, of litter, and of bare ground either did
not differ among vegetation types or were signifi­
cantly higher in successional prairie.
Live shoot mass was significantly higher in Agro­
pyron-dominated vegetation than successional prairie
(Table 2). Total root mass was 3-5 times greater than
live shoot mass but did not differ between vegetation
types, nor did root: shoot ratio.
Discussion
Competition
The results clearly indicate that competition from
neighboring vegetation reduced the growth rate and
increased biomass allocation to below ground struc­
tures in both transplant species. These responses are
typical of species competing for belowground re­
sources (Casper and Jackson 1997).
122
Successional prairie
Agropyron-<lominated
vegetation
08
§
~ 0.6 ~
I?
0.8 ,
1
0.5
07 ~
AQItlPY"'"
0.71
Y =0,26 -0.98.
R' =0 .37
~
0 .4
c:
C1.
~ 0.2
0.1
0.3
..'. ..
0.2
0.1
o +-~---r-"'~--T~
o
0.1 0.2
0.3 0.4
Boutelous
y = 0.23 • 0 .69 •
R.1_0. 15
01
'i:: 0.4
~ 0.3
~ 0.2
~
0.1
o
0.1
0.2
0.3 0.4
0.5
0.8
0.7
~ 0.6
{:! 0 .5
o
0.5
0.8
§
R' =0 .40
0.4
.!2 0.3 •
~
Agropyron
y' 0.32· 0.86.
0.6
0 .5
BouIIlioue .
0.7
Y • 0.17 • 0.42 •
0.6
0.5
R' = 0.21
0.4
0.3
~
. ..
a +-~..:...,-=-~~~
o 0.1 0.2 0.3 0.-4 0.5
0.2
1
oi ~.
0.1
0.1
0.2
0.3
0.4
0.5
Log (aboveground nel~hbor mass + 1)
(g lube' )
Figure 5. Growth of Agropyron and Bouteloua transplants as a
function of neighbor shoot mass in all neighbor (AN) tubes in
Agropyron-<lominated vegetation and successional prairie. Data are
combined across levels of water availability. Lines indicate signif­
icant (P ~ 0.05) regressions. Mean actual growth in no neighbor
(NN) tubes, shown by the open symbol on the y-axis of each graph,
was not included in the regres sions.
Competitive responses varied among vegetation
types. In the successional prairie, competition re­
duced the growth rate of Bouteloua by about 60% and
that of Agropyron by about a third (Figure 3, right),
indicating that Agropyron had a stronger ability to re­
sist competitive suppression. In Goldberg (1990) ter­
minology, Agropyron is a relatively strong response
competitor in this vegetation type. The spread of
Agropyron into native prairie (Coffin et al. 1996) may
relate to its ability to resist competitive suppression.
In contrast, competition in the Agropyron-dominated
vegetation reduced the growth of both species by
about half (Figure 3, left), indicating that both spe­
cies were suppressed by about the same extent Simi­
lar result s have been reported when Agropyron and
Bouteloua were grown with neighboring Bromus in­
ennis, another introduced C 3 grass (Gerry and Wil­
son 1995). Thus, the ability of Agropyron to resist
competitive suppression does not explain its domi-'
nance in fields of introduced grasses (Table 2).
Competitive effects may prevent other species
from establishing in Agropyron-dominated vegeta­
tion. The F-ratio of the competition term (Table I)
was twice as large in Agropyron-dominated vegeta­
tion as in successional prairie, suggesting that Agro­
pyron-dominated vegetatio n has strong competitive
effects and that competition accounts for more varia­
tion in transplant performance in this vegetation type
(Underwood and Petrait is 1993). Strong competitive
effects would explain why Centaurea diffusa Lam.,
an introduced herb, invaded plots of Agropyron at a
much lower rate than it invaded unseeded plots or
plots of Elymus junceus Fisch. (Berube and Myers
1982).
In the absence of neighbors, Bouteloua transplants
grew larger (Figure 5) and faster (Figure 3) than
Agropyron transplants. The large difference between
predicted and actual growth of Bouteloua in the ab­
sence of neighbors (Figure 5), however, indicates that
even low amounts of neighbors greatly reduced the
growth of Bouteloua. The first few neighbors appear
to have a disproportionately large competitive effect
on Bouteloua. In contrast, the predicted and actual
growth of Agropyron in the absence of neighbors
were very similar (Figure 5), suggesting that all
neighbors had a proportionate effect on Agropyron.
This result supports the conclusion that Agropyron
was less suppressed by neighbors and therefore is a
stronger response competitor than Bouteloua.
Growth was strongly affected by factors other than
competition, since aboveground neighbor mass ex­
plained no more than 40% of the variation in trans­
plant growth (Figure 5). Environmental heterogeneity
may contribute to the variation in growth: species
composition and amount of bare ground varied
among tubes. Other factors that could affect trans­
plant growth include genotype, initial mass, and her­
bivory (Goldberg 1987).
Species identity
Survival and root: shoot ratio were strongly affected
by transplant species identity (Table 1). Interspecific
differences reflect genetic variation but may also re­
flect biotic factors such as species-specific herbivory.
More insect herbivory was observed on Agropyron
than Bouteloua transplants (1. Bakker, personal obser­
vation), which may have contributed to the lower sur­
vival (Figure 2) and higher root: shoot ratios (Fig­
ure 4) of Agropyron. Similarly, the presence of neigh­
bors may have protected transplants from herbivores
(Wilson 1994), explaining the tendency for higher
survival of Agropyron transplants in the presence of
123
Table 2. Spec ies richne ss. cover. and biomas s (rneane.Sli) of Agropyron-dominaled vegetation and successional prairie in Grasslands Na­
tional Park. Saskatchewan. Canada. n
=8 per vegetation type .
Attribute
Agropyron·dominated vegetation
Spe cies richness (spec ies per 0.5 m')
6.4:1:0.7
x'
1O.1±O.6
8.25··
Co ver (%)
Bare ground
30.7±6.6
17.6:1:4.7
2,09
Liner
7.5:1:2.2
7 .5±2 .2
Agropyron cristatum (L.) Gaertn,
31.9±3.7
0.4±0.4
0 .00
13.06··'
Artemisia frigida Willd .
14.3±7.3
6.0±2.0
0.12
Bouteloua gracilis (HBK.) Lag .
3.9±2.4
4.2±2.4
0.03
Lichens
3.2±1.7
2 .7±0 .3
1.20
Stipa comata Trin, & Rupr,
0.2±O.1
1O.9±7.6
Koeleria crlstata (L. ) Pers,
o.rso.:
18.8±4.5
6.67"
9 .86··
Selaginella densa Rydb .
0.1±0.1
7 .5±2 .8
2 .14
Pascopyrum smithii (Rydb.) Love
O.O±O.O
4 .9±2.2
8.47··
Poa sandbergii Vasey
O.O±O.O
1.6±O.5
6.67·'
Sbool mass (g m-')
80.9±11.2
52.8±3 .9
4.41·
Root rna." (g rn-')
Root sboot ratio
240 .1:1:31.2
248 .3±31.5
0.04
3.3±0.6
5.2:1:1.1
1.33
neighbors in the Agropyron-dominated vegetation
(Figure 2).
Over the course of the experiment, Agropyron al­
located slightly more biomass to roots than shoots
while Bouteloua allocated almost twice as much bio­
mass to shoots as to roots. These differences in bio­
mass allocation likely reflect ontogenetic patterns, as
C 3 species prepare for dry, warm summer conditions
by building up belowground reserves (Morgan et al.
1998).
Water availability
The absence of significant effects of water availabil­
ity may relate to the timing of pulses in relation to
plant phenological stage (Bilbrough and Caldwell
1997). Factors that control one stage in a plant's life
history may not be critical at another stage. For ex­
ample, seed germination under water stress and seed­
ling survival under water stress are not correlated in
several C 3 and C 4 grasses (Qi and Redmann 1993).
In this experiment, water availability did not affect
the survival, growth, or biomass allocation patterns of
older transplants even though soil water availability
strongly controls the germination and establishment
of both transplant species (Briske and Wilson 1980;
Qi and Redmann 1993; Lauenroth et al. 1994). Boute­
loua has a high water use efficiency (Morgan et al.
1998) and is physiologically affected by small water
events (Sala and Lauenroth 1982), but the growth of
individual plants does not appear to be limited by
water availability.
Also, processes during the interpulse interval
(Goldberg and Novoplansky 1997) may have pre­
vented the detection of significant effects of water
availability. Interpulse processes include transplant
moisture stress and mortality.
Although community-level production in the
northern Great Plains is strongly correlated with pre­
cipitation (Sala et al. 1988), water availability was not
a limiting factor in the growth or biomass allocation
patterns of Agropyron or Bouteloua. Other resources
such as nitrogen may have limited transplant growth
(Casper and Jackson 1997). A study in North Dakota
found that water availability limited Pascopyrum
smithii (Rydb .) LOve biomass production and nitro­
gen availability limited Agropyron biomass produc­
tion (Frank and Ries 1990). Interspecific differences
in resource limitation help explain species coexist­
ence and the paradoxical result that water limits com­
munity-level production but does not limit the growth
of a particular species.
Vegetation types
The presence of Bouteloua in both vegetation types
(Table 2) suggests that individuals either survived
cultivation or invaded from adjacent prairie after cul­
tivation ended. In light of the strong competitive ef­
fects of Agropyron (Table 1) and the fact that compe­
124
tition from existing vegetation greatly reduces the es­
tablishment of Bouteloua (Samuel and Hart 1992;
Aguilera and Lauenroth 1995), it seems probabl e that
Bouteloua establishment occurred before Agropyron
was sown. Cover of Bouteloua was much lower in ei­
ther vegetation type (4%, Table 2) than in unbroken
prairie nearby (15%, Christian and Wilson (1999)),
illustrating its slow recovery after disturbance (Sam­
uel and Hart 1994; Coffin et at. 1996).
The community-level characteristics reported by
this study (Table 2) correspond with those of many
other studies (Hull and Klomp 1967; Smoliak et aI.
1967; Looman and Heinrichs 1973; Redente et aI.
1989). For example, aboveground production is
higher for Agropyron than for native species
(Lawrence and Ratzlaff 1989) and below ground pro­
duction is lower (Smoliak et al. 1967) or equal (Re­
dente et at. 1989) than in successional prairie or un­
broken prairie . In addition, soil organic mailer levels
are reduced in Agropyron-dominated vegetation (Dor­
maar et at. 1995). Together, these results suggest that
Agropyron exports nutrients from the soil (Dormaar
et at. 1995; Lesica and DeLuca 1996) and has signif­
icant long-term ecosystem effects (Christian and Wil­
son 1999). Competitive ability is a significant factor
contributing to the maintenance of existing Agropy­
ron populations and the invasion of new areas by
Agropyron. Therefore, Agropyron will continue to af­
fect ecosystem structure and function in the northern
Great Plains for the foreseeable future .
Acknowledgements
We thank L. Ambrose, E. Bakker, J. Christian, M.
Kochy, D. Peltzer, and T. Willow for field assistance,
Grasslands National Park for lodging, and B. Foster,
D. Bakker, and E. Bakker for improving earlier drafts
of this work. TIlls paper is based on work done in
partial fulfillment of the requirements for the M.Sc .
degree at the University of Regina and was supported
by Grasslands National Park, the Canada­
Saskatchewan Agricultural Green Plan Agreement
and the Natural Sciences and Engineering Research
Council of Canada.
References
Aguilera M .O. and Lauenroth W.K. 1995. Influence of gap d isturb­
ance s and type of rnicrosites on seedli ng establishment in
Bouteloua gracilis . Journal of Ecology. 83 : 87- 97.
Barkwonh ME. and Dewey DR. 1985. GenomicaJly based genera
in the perennial Triticaceae of North America : Identification
and membership. American Journal of Botany. 72 : 767-776.
Berube D.E. and Myers J.H. 1982. Suppression of knapweed inva­
sion by crested wheatgrass in the dry interior of Briti sh Colum­
bia. Journal of Range Management . 35: 459-46 I.
Bilbrough CJ . and Caldwell M.M. 1997. Expl oitation of spring­
time ephemeral N pulses by six Great Basin plant species. Ecol­
ogy. 78: 231-243.
Briske D.O. and Wilson A .M. 1980. Drought effects on adventi­
tious root development in blue grama seedlings. Journal of
Range Management. 33: 323-327.
Casper B.B. and Jackson R.B. 1997. Plant competition under­
ground. Annual Review of Ecology and Systematics. 28: 545­
570.
Christian J.M . and Wilson S.D . 1999. Long-term ecosystem im­
pact s of an introduced grass in the northern Great Plains. Ecol­
ogy . 80: 2397-2407.
Coffin D.P., Lauenroth W.K. and Burke I.C, 1996. Recovery of
vegetation in a semiarid grassland 53 yea rs after disturbance.
Ecological Applications. 6: 538-555.
0' Antonio C.M. and Vitousek P.M. 1992. Biological invasions by
exotic gm.sses, the grasslfire cycle, and global change. Annual
Re view of Ecology and Systematics. 23 : 63-87.
Donnaar J.E , Nae!b M.A.. Willms W.O. and Chanasyk D.S. 1995.
Effect of native prairie, crested wheatgrass (Agrop yron cris ­
tatum (L.) Gaertn .) and Russian wildrye (Elymus [unceus
Fisch .) on soil chemical properties. Journal of Range Manage­
ment . 48: 258-263.
Drake J.A. , Mooney H.A ., di Castri F.• Groves R.H ., Kruger FJ.,
Rejrnanek M. et al. (eds) 1989. Biological invas ions: a global
perspective. John Wiley, New York, New York.
1966-1993. Monthly record : meteorological o bservations in Can­
ada . Minister of Suppl y and Services Canada, Ottawa, Ontario.
1993 . Canadian climate normals, 1961-1990. Minister of Supply
and Services Canada, Ottawa, Ontario.
Frank A .B. 1994. Physiological comparisons of crested wheatgrass
and western wheatgrass to warer, Journal of Range Manage­
menl. 47: 460-466.
Fronk A .B. and Ries R.E. 1990. Effect of soil water, nitrogen, and
growing degree-days on morphological development of crested
and western wheat grass. Journal of Range Management. 43 :
257-260.
Gerry A.K. and Wilson SD. 1995. The influence of initial size on
the com petitive responses of six plant species. Ecology. 76 :
272-279.
Goldberg D.E. 1987. Neighborhood competition in an old -field
plant community. Ecology. 68: 1211-1223.
Goldberg D.E. 1990. Components of resource competition in plant
communities. In : Gra ce J.B . and Tilman D. (eds), Perspectives
On Plant Competition . Academic Pres s, San Diego, California,
pp.27-49.
Goldberg D. and Novoplansky A. 1997. On the relative importance
of competition in unproductive environments. Journal of Ecol­
ogy. 85: 409-418.
125
Hull A.C. and Klomp GJ . 1967.1bickening and spread of crested
wheatgrass stands on southern Idaho ranges . Journal of Range
Management. 20: 222-227 .
Lauenroth W.K., Sala O.E., Coffin D.P. and Kirchner T.B. 1994.
The importance of soil water in the recruitment of BOUIe/DuD
gracilis in the sbortgrass steppe. Ecological Applicat ions. 4:
741-749.
Lawrence T. and Ratzlaff C.D. 1989. Performance of some native
and introduced grasses in a semiarid region of western Canada.
Canadian Journal of Plant Science. 69: 251-254.
Lesica P. and DeLuca T.H. 1996. Long-term harmful effects of
crested wheatgrass on Great Plains grassland ecosystems . Jour­
nal of Soil and Water Conservation. 51: 408-409.
Looman J. and Best K.B. 1987. Budd's 1I0ra of the Canadian prai­
rie provinces. Publication 1662, Agriculture Canada Research
Branch.
Looman P.E. and Heinrichs D.H. 1973. Stability of crested wheat­
grass pastures under long-term pasture use. Canadian Journal
of Plant Science. 53: 501-506.
Morgan J.A.. LeCain D.R., Read 1.1., Hunt H.W. and Knight W.G.
1998. Photosynthetic pathway and ontogeny affect water rela­
tions and the impact of CO 2 on Bouteloua gracilis (C.,) and
Pascopyrum smithii (C.,). Oecologia . 114: 483-493.
Mueller-Dombois D. and Ellenberg H. 1974. Aims and Methods of
Vegetation Ecology. John Wiley, New York, New York.
Qi M .Q. and Redmann R.E. 1993. Seed germination and seedling
survival of C, and C. grasses under water stress. Journal of
Arid Environments . 24: 277-285.
Redenle E.F.. Biondini M.E. and Moore J.C. 1989. Observations
on biomass dynamics of a crested wheatgrass and native short­
grass ecosystem in southern Wyoming. Journal of Range Man­
agement. 42: 113-118.
Sala O.E. and Lauenroth W.K. 1982. Small rainfall events: an eco­
logical role in semiarid regions. Oecologica. 53: 301-304.
Sala O.E., Parton WJ., Joyce L.A. and Lauenroth W.K. 1988. Pri­
mary production of the central grassland region of the United
States. Ecology. 69: 40-45.
Samuel MJ . and Hart R.H. 1992. Surv ival and growth of blue
grama seedli ngs in competition with western wheatgrass. Jour­
nal of Range Managemenl. 45: 444-448.
Samuel MJ . and Hart R.H. 1994. Sixty-one years of secondary
succession on rangelands of the Wyom ing High Plains . Journal
of Range Management. 47 : 184-191.
1994. JMP Statistics and Graphics Guide, version 3. SAS Institute,
Cary, North Carolina.
Srnoliak S., Johnston A. and Lurwick L.E . 1967. Productiv ity and
durability of crested wheatgrass in southeastern Alberta , Cana­
dian Journal of Plant Science. 47 : 539-548.
Sutler G.C. and Brigham R.M. 1998. Avifaunal and habitat changes
resulting from conversion of native prairie to crested wheat
grass : patterns at songbird community and species levels. Can.
J. Zoo!. 76: 869-875 .
Underwood A.J . and Petra itis P.S. 1993. Structure of intertidal as­
semblages in different locations: how can local processes be
compared. In: Ricklefs R.E. and Schluter D. (eds), Species Di­
versity in Ecological Communities. University of Chicago
Press, Chicago, Illinois.
Vitousek P.M. 1990. Biological invasions and ecosystem processes:
towards an integration of population biology and ecosystem
stud ies. Oikos, 57: 7-13.
Wilson S.D. 1989. The suppression of native prairie by alien spe­
cies introduced for revegetation. Landscape and Urban Plan­
ning. 17: 1I3-119.
Wilson S.D. 1994. The contribution of grazing to plant diversity in
alpine grassland and heath . Australian Journal of Ecology. 19:
137-140.
Wilson S.D. and Belcher J.W. 1989. Plant and bird communities of
native prairie and introduced Eurasian vegetation in Manitoba.
Canada . Conservation Biology. 3: 39-44.